Author: Robert Hyatt
Date: 11:21:03 10/31/99
Go up one level in this thread
On October 31, 1999 at 12:26:22, Bruce Moreland wrote: >On October 31, 1999 at 09:38:30, Robert Hyatt wrote: > >>I don't follow. Murray defined R=2 and recursive null-move. I simply didn't >>like them early on. I tried them when I first started Crafty, as my intent was >>to try _everything_ again (using all my Cray Blitz 'notes', plus anything else >>I saw in the literature.) IE I looked at Gnuchess, and it was doing null-move >>R=2 recursive well prior to 95 when I started Crafty. That bumped it up in >>priority for me. >> >>In any case, I consider it far more important to publish an idea, than to later >>publish details about implementing an idea someone else had already tested. >>Chrilly didn't modify the original algorithm for null-move at all. His main >>contribution (IMHO) was the idea of useing the null-move result to detect >>threats. But 'null-move' itself was 'prior work'. > >I think you started doing R=2 after talking to me, and I did R=2 based upon the >Donninger article and was enthusiastic about the results. If you think back, when you suggested that, remember my first comment? "R-2 is _bad_" :) I had tried that in Cray Blitz, because Murray suggested in his paper that "R=2 and up needs further study..." But, if you remember further, I was running on a P5/133, and not going very fast early on (about 5K nodes per second early in the bitmap stuff as I did copy/make). And you kept mentioning R=2, and I tried it/disabled it multiple times (those interested can find the repeated tests in main.c comments where it was on, off, on, off, etc. :)) But my first run-in with R=2 was Murray's experiments in the late 80's... It just turned out that in the late 80's, on the machine I was testing on, R=2 also looked bad because of the 5-6 plies we could do on non-cray boxes. Murray came to the same conclusion for the same reason, if I remember correctly. > >I don't understand the point of this thread. It is clear that if you want to >learn good stuff, you read articles on research programs or talk to amateurs. >That's the meat and potatoes, which is not subsumed by the small amount of gravy >you'll get from the professionals. I wouldn't disagree at all. Particularly if past history is used to predict the future. :) But there was a time when the best in the world were freely explaining what we were doing, including (the ACM dominants from 1970-1994) Slate, then Thompson, then myself, then Hsu/Campbell. > >I don't know why Donninger is being used as an example of a professional. He is >professional now, sure, but in 1993 he was amateur. He was technically amateur >as late as 1995 and possibly longer. That has been mentioned, but seems unimportant to others. > >I have benefited with discussion with amateurs primarily, and from reading >articles in the ACC books, a few other books, and the ICCAJ's. > >I think the amateurs figured out the internet before the professionals, so >that's at least some of this. You'll talk to whoever is there, and so far it's >been amateurs. > >I have received a few ideas from professionals. David Kittinger in particular >provided me with some help at the 1995 WCCC. I've got a few ideas from Ed's >posts. Frans told me something at a WMCCC that I couldn't make work. I have >compared notes with Stefan and John Stanback. > >Perhaps the professionals have gotten ideas from me, I don't know. > >I don't think that there is much magic in computer chess programs. The >professionals aren't doing anything special, they've just tuned for a long time. > If you talk to them, if you are wimpy you'll get a lot of info, but you could >get this from amateurs. Once your program gets stronger you get ideas from >frank discussions, but the ideas don't always work in your implementation. > >bruce
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.